US 6964936 B1
A method of making a catalyst with monolayer or sub-monolayer metal by controlling the wetting characteristics on the support surface and increasing the adhesion between the catalytic metal and an oxide layer. There are two methods that have been demonstrated by experiment and supported by theory. In the first method, which is useful for noble metals as well as others, a negatively-charged species is introduced to the surface of a support in sub-ML coverage. The layer-by-layer growth of metal deposited onto the oxide surface is promoted because the adhesion strength of the metal-oxide interface is increased. This method can also be used to achieve nanoislands of metal upon sub-ML deposition. The negatively-charged species can either be deposited onto the oxide surface or a compound can be deposited that dissociates on, or reacts with, the surface to form the negatively-charged species. The deposited metal adatoms can thereby bond laterally to the negatively-charged species as well as vertically to the oxide surface. Thus the negatively-charged species serve as anchors for the metal. In the second method, a chemical reaction that occurs when most metals are deposited on a fully hydroxylated oxide surface is used to create cationic metal species that bind strongly both to the substrate and to metallic metal atoms. These are incorporated into the top layer of the substrate and bind strongly both to the substrate and to metallic metal atoms. In this case, these oxidized metal atoms serve as the anchors. Here, as in the previous method, nanoislands of catalytic metal can be achieved to increase catalytic activity, or monolayers or bilayers of reactive metal can also be made.
1. A method for making a heterogeneous catalyst, comprising:
cleaning the surface of a support material;
introducing a negatively-charged species onto said surface of said support material; and
depositing a catalytic metal on said surface in a closed environment under vacuum conditions at less than two monolayers of coverage of said catalytic metal.
2. The method of
3. The method of
4. The method of
5. The method of
6. The method of
7. The method of
8. The method of
9. The method of
10. The method of
11. The method of
12. The method of
13. The method of
14. The method of
15. The method of
16. The method of
17. The method of
18. The method of
19. The method of
20. The method of
21. The method of
This invention was made with Government support under Contract No. DE-AC04-94AL85000 awarded by the Department of Energy. The Government has certain rights in the invention.
This application is related to U.S. patent application Ser. No. 09/576,919, filed on May 23, 2000.
The invention describes a method for making a heterogeneous catalyst and, more particularly, to a method for making a metal heterogeneous catalyst with monolayer (ML) or sub-ML metal thickness.
Catalyst technology plays a critical role in the production of materials related to many facets of the world economy including petroleum refining, pharmaceutical productions, chemical processing and production, and environmental cleanup. Heterogeneous catalytic reactions are widely used and are commonly characterized by reactions performed with the reactant(s) and product(s) in the fluid or gas phase and the catalyst in the solid phase. In heterogeneous catalytic reactions, the reaction occurs at the interface between phases; that is, the interface between the fluid or gas phase of the reactant(s) and product(s) and the solid phase of the supported catalyst. Hence, the properties of the surface of a heterogeneous supported catalyst are significant factors in the effective use of that catalyst. Specifically, the surface area of the active catalyst, as supported typically on a metal oxide substrate, and the accessibility of that surface area to reactant chemisorption and product desorption are important. These factors affect the activity of the catalyst, defined by the rate of conversion of reactants to products. The chemical purity of the catalyst and the catalyst support have important effects on the selectivity of the catalyst, which is the degree to which the catalyst produces one product from among several products, and the lifetime of the catalyst.
Most heterogeneous catalysts are composed of a selected combination of active material, promoter and support. Catalysts with high surface areas are desirable to reduce the cost of material and to increase the activity of the catalyst (that is, the product production rate per unit weight). Active catalytic materials for most non-biological chemical reactions come generally from the transition metals of the Periodic Table and are generally considered to include vanadium, iron, cobalt, nickel, molybdenum, ruthenium, rhodium, palladium, cadmium, tungsten, rhenium, osmium, iridium, platinum and mercury, although other metals are catalytic in specific reactions. Promoters are included to increase activity and stability and use compounds, particularly oxides, which include elements selected from lithium, sodium, magnesium, boron, potassium, calcium, barium, lanthanum, cerium, and thorium. The supports are typically chosen from materials that have high surface areas and may or may not contain sites active for catalyzing specific reactions and generally include Group IIa and IIIa elements, alkaline earth and transition metal oxides (such as Al2O3, SiO2, TiO2, and MgO, zeolites and activated carbon). Low surface area supports, or substrates, can also be used. These substrates include ceramics, such as the high temperature form of alumina, or metallic monoliths.
Generally, catalytic activity is proportional to catalyst active surface area. Therefore, a high specific area is desirable. That surface area must be accessible to reactants and products as well as to heat flow. The chemisorption of a reactant by a catalyst surface may be preceded by the diffusion of that reactant through the internal structure of the catalyst. To minimize cost and maximize catalytic activity, it would be desirable for metal catalytic materials to be deposited on supports at levels of sub-MLs to a few MLs in thickness. When the catalytic material is a metal and the support is an oxide, the metal deposited on the oxide surface generally does not wet the surface because the metal-to-oxide interaction is relatively weak. At room temperature, metal islands form on an oxide. Because it is known that reactions often occur at step sites, these islands, having many edges and steps, are desirable in this sense, but all interior metal material is in effect wasted. Thus a reduction in island size is highly desirable.
The morphology of the resulting metal layer is thus quite variable within several nanometers (nm) of the oxide surface, and the resulting interface is typically quite non-uniform and weak. To strengthen the interface, reactive metals are sometimes added, such as brazing compounds. However, it is sometimes undesirable for such metals to be added because they often react with the oxide to form an intermediate layer, which is poorly defined and contrary to the desired catalytic structure. Additionally, the shape of the particles on the support surface depends on the interfacial energy between the metal and the oxide, and can in principle vary from flat two-dimensional islands (if strong interactions are present) to three-dimensional amorphous or faceted objects having minimal contact to the oxide (if the interaction is very weak). The ability to control this shape by interfacial engineering of the adhesion energy would provide an additional tool for catalyst design.
Useful would be a method for producing a catalyst with maximally dispersed and controlled ML or sub-ML amounts of active material on a support in order to produce a more effective and more cost efficient catalytic material.
A significant expense in catalyst manufacture is the cost of the precious metals needed. To improve efficiency, catalysts are often made with small three-dimensional (3D) metal particles, thereby achieving a high surface area per metal weight. If the method of manufacture involves the deposition of a metal (the catalyst) on an oxide surface (the support), normally, the metals grow as 3D islands. If the deposition is slow and at moderate temperature, the size of these islands can be remarkably uniform.
In the method of the present invention, anchors are made on or within the substrate surface to control the morphology of deposited reactive metal. In the first embodiment of the present invention, a catalyst is prepared comprising a metal material and a support material with near total surface hydroxylation to permit the key chemical reaction that achieves anchor formation and nanometer island stability. A promoter material can optionally be added. The catalytic metal is deposited on the hydroxylated support in a closed environment under vacuum conditions to form 2D (two-dimensional) laminar layers (
It is important to note that ultrathin metal films on an oxide support have been shown to have enhanced catalytic rates compared with thicker films for certain reactions, and that these rates can be about the same as those of nanoclusters. Thus, depending on the specific reaction and catalyst stability, one or two ML layers can be as or more useful than 3D nanoislands.
In the second embodiment of the present invention, an oxide surface is used that is only partially hydroxylated. This can be achieved for example by heating fully hydroxylated alumina (which is in fact any alumina that has been exposed to ambient conditions) to about 1000 K. In this case, the isolated hydroxyl groups now serve as the anchors and metal nanoislands grow around them at sub-ML depositions. In this case, noble metals as well as others will make nanoislands on an otherwise clean alumina surface (
Thus, the two embodiments differ in the nature of the anchors: metal cations in the first, and nearly isolated negatively-charged species, such as hydroxyl ions, in the second. They also differ in the treatment during manufacture: room-temperature conditions in the first, but high temperature annealing may be required in the second, unless the negatively-charged species can be deposited directly from other sources, as described below.
It is sometimes also possible to make useful structures that are metastable in that they are not energetically the globally preferred structures. For example, the stability of metal islands is related to the energy required to detach atoms from their edges on the surface. This sets a temperature for island growth on the surface (Ostwald ripening). Thus if this diffusion occurs at a given temperature, the metal adatoms move until they join a growing metal island. However, if the strength of the adatom binding could be increased to the point at which the adatom is pinned and cannot move at the manufacturing temperature, then the deposited metal will remain metastable, again producing a more uniformly dispersed surface (illustrated in
In the method of the present invention, a catalytic metal is deposited on a support by controlling the wetting characteristics of the support surface. In one embodiment, a fully hydroxylated surface is used to promote the chemical reaction (Eq. 1) described previously. In another embodiment, a sub-ML of negatively-charged species, such as hydroxyl groups, is produced not by high temperature heating but is introduced to the support surface. Thereby a metal either wets the surface in a controlled and uniform manner in a growth mode that promotes layered surfaces, or, with sub-ML deposition, forms nanoislands, thus achieving maximal dispersion of the expensive reactive metal. This also produces the greatest yield of the product per surface area. The supports are typically chosen from materials that have surfaces containing sites active for catalyzing specific reactions and generally include Group IIa and IIIa elements (such as materials comprising the elements boron, aluminum, gallium, and indium), alkaline earth materials (including materials comprising the elements beryllium, magnesium, calcium, strontium, and barium) and transition metal oxides (such as Al2O3, TiO2, SiO2, and MgO). Low surface area supports, or substrates, can also be used. These substrates include ceramics, such as alumina, or metallic monoliths.
Introduction of a negatively charged species to the surface can be accomplished by either depositing a negatively-charged species onto the oxide surface or by depositing a compound onto the surface that can dissociate or otherwise react to form a negatively-charged species. The adhesion or binding of the metal atoms to the surface is increased because the deposited metal ions can bond laterally to the negatively charged species as well as vertically to the oxide surface in a layer-by-layer metal growth mode.
Interestingly, the adhesion mechanism changes with the fraction of coverage of the negatively charged species on the surface of the support, such as a metal oxide support. With less than approximately ½ ML coverage of the negatively-charged species, the primary mechanism appears to be that the adjacent metal adatoms experience electrostatic interactions both vertically with the underlying O2− ions and laterally to the introduced negatively charged species, resulting in a deposited layer of greater uniformity and greater adhesion. However, when the surface is essentially fully covered, such as an essentially fully hydroxylated surface, certain metals produce a chemical reaction that leads to laminar growth of the metal film or the production of metal nanoislands, where the metal cations are incorporated into the top layer of the metal oxide (
One useful negatively charged species is an adsorbed isolated hydroxyl group. These can be added to a clean and largely defect free oxide surface by means such as exposure of the surface to water (if the oxide readily dissociates water) and subsequent heating to remove a portion of them, to a water plasma, to a mixture of oxygen and water gas, to an oxygen plasma in the presence of a background of hydrogen gas, or to hydroxide-containing chemicals which readily dissociate on the surface leaving a hydroxyl group behind (such as specially made chemicals as XeOH).
Another useful negatively charged species is an adsorbed isolated oxygen atom. These can be added to the surface by exposure to and slow cooling in an oxygen plasma, by exposure to chemicals which dissociate on the surface leaving an oxygen atom behind, such as ozone or, in the presence of ultraviolet light, nitrous oxide (N2O). Other useful negatively charged species are halogens such as fluorine, chlorine, bromine, and iodine, deposited, for example, by XeF or other like compounds.
In one embodiment, the negatively charged species is at a coverage of less than approximately one-half ML; theoretical calculations have indicated that a cover of one-third ML is optimal. To achieve approximately one-third ML OH coverage, a surface is first exposed to humid air sufficient to achieve essentially one ML coverage and then the surface is heated to approximately 1000–1200K, although lower temperatures may also work, to reduce the hydroxyl coverage to approximately one-third ML. Coverage as low as one-tenth ML or less can be achieved by controlling the deposition time in conjunction with the amount or concentration of reactant metal made available at the surface. Essentially any metal adatom can be deposited on an oxide surface for this embodiment. The requirement is that the atom be a positively charged adatom or bind sufficiently strongly to the site adjacent to the negatively-charged species that nanoislands or layer-by-layer growth can occur. Metal adatoms in Groups IB, IIB, IIIB, IVB, VB, VIB, VIIB, and VIII of the Periodic Table of Elements all satisfy these criteria.
In particular, the metal adatoms lithium, cobalt, potassium, yttrium, niobium, ruthenium, palladium, platinum, copper, silver, gold and aluminum have been investigated theoretically using the Born-Haber cycle calculations and have been shown to wet under the method of the present invention and to have increased adhesion. Copper (one of these adatoms) deposition on a ceramic surface, such as an alumina surface, is of particular interest because of its use in microelectronics; experimental evidence has validated the theoretical calculations that show the increased adhesion of copper to the oxide layer through the addition of a negatively-charged species according to the method of the present invention (
Metal adatoms can be deposited in this embodiment to promote nanoisland or layer-by-layer metal growth on a surface. More than one metal can be deposited by combining the metals or by combining a metal and a promoter material. Promoters are included to increase activity and stability and include compounds, particularly oxides, which include elements selected from lithium, sodium, magnesium, boron, potassium, calcium, barium, lanthanum, cerium, and thorium. One surface of interest includes aluminum oxide, as films of this material are easily produced on metals such as cobalt by Al metal deposition and oxidation at moderate temperatures. Aluminum oxide is also an important support material for heterogeneous catalysts and is an important structural ceramic. Another oxide surface of interest includes silicon dioxide (SiO2) surfaces. On such a surface it may be necessary to sputter the surface to increase the density of available binding sites. Both metal oxide surfaces and non-metal oxide surfaces, such as silica oxide surfaces, present oxygen sites to which the positively-charged adatoms can bind and positively charged sites, such as from the aluminum and silicon ions, to which the added negatively-charged species can bind to promote the layer-by-layer growth mechanism. Thus, the methods described herein can apply to all of the numerous metal/oxide junctions.
The isolated adatoms are positively charged because they coordinate to (i.e. are adjacent to) negative ions on the surface. If the oxide surface is prepared to also have negatively charged adspecies upon it, these species attract metal adatoms and increase their binding (see
At the correct concentrations, these negatively charged species could dramatically increase the binding strength of the resulting metal overlayer, thus adding both stability and smoothness to the interfacial properties. The specific identity of the negatively charged species will affect the strength of the effect and other characteristics, such as whether products will be evolved by dissociation of the species in the presence of the metal overlayer. Because the negatively charged species repel each other, saturation coverage of the species excludes them from being on adjacent surface sites (
In one embodiment, the surface is cleaned to remove impurities. These methods include, for example, sonicating in a solvent, sputtering removal using an ion beam, burning off impurities by using an oxygen plasma or an ozone source, or annealing at high temperature in an oxygen ambient atmosphere.
The quality of the surface with respect to defects in the crystal structure or non-stoichiometries in the oxide composition can also be improved. This can be done, for example, by annealing at high temperature in an oxygen environment. This step can be combined with the cleaning step.
Negatively charged species can then be introduced to the surface. This can be accomplished, for example, by exposure of the surface to atoms or molecules, or to mixtures of atoms/molecules, at appropriate temperatures and pressures, so as to cause attachment to, or in the case of molecules spontaneous dissociation on, the surface, so as to produce and stabilize negatively-charged fragments on the surface, or to react with the surface to convert the surface into a surface with negatively-charged species on the surface, and so that negatively charged fragments remain after treatment. This introduction of negatively-charged species to a surface can be accomplished by gases reacting with the surface, which converts the surface into one with negatively-charged species on top of a close packed plane; by stimulating the dissociation of molecules on the surface using photon (for example, ultraviolet) or electron beams; or by exposure of the surface to a plasma which leaves such species on the surface at processing temperatures. The negatively charged species can be introduced at room temperature or at elevated temperatures. The metal is then deposited on the surface, wetting the surface in a controlled manner. In one embodiment, the first metal layer could be introduced simultaneously with the negatively charged species, wetting the surface in a controlled manner. An additional step of annealing the surface can be performed to provide additional stability to the interface system, as metastable interfaces have been observed to form with mixed neutral and positively charged metal atoms.
In one embodiment using hydroxyl ions as the negatively charged species introduced to the surface, when the surface cover of the hydroxyl ions is essentially complete (with essentially complete being greater than approximately 0.9 ML coverage), a chemical reaction (reaction (1) above) results from deposition of certain metals that results in metal cations (
When clean aluminum oxide is exposed to ambient conditions or approximately 1 Torr or more of water, the surface changes to one terminated entirely by OH. The surface is robust, requiring high temperature (up to 1000K) to remove sufficient OH to achieve a ⅓ ML coverage of OH. When using essentially complete OH coverage, high temperature heating is not required however. The metal oxide surface can simply be first cleaned and the metal atoms of interest deposited. The resulting deposition yields a metal strongly adhered to the oxide surface. For example, aluminum oxide was first cleaned, such as by sonication in a solvent such as acetone and isopropanol, and then in vacuum by oxygen plasma cleaning at room temperature, giving a surface essentially free of carbon with trace amounts (less than 0.1 ML) of metals such as Mg and Ca (which are probably subsurface). Co depositions were carried out at ambient temperature using an effusion cell. Analyses showed that a chemical reaction occurred to produce the cationic Co anchors,
In another embodiment wherein the negatively charged species (in this case, hydroxyl ions) is at a lower fractional coverage, copper is wetted onto a cleaned alumina surface.
The experimental data confirm the theoretical predictions. A sapphire sample was first cleaned by sonication in an aqueous solvent, such as acetone, methanol and deionized water prior to insertion into a vacuum system, with working pressures in the analysis chamber of approximately 1–5×10−9 Torr and in the range of approximately 10−8 to 10−7 Torr in the sputter deposition chamber. Hydroxyl ions were introduced onto the sapphire surface by exposure to ambient air at moderate humidity, and then heated to approximately 1000K, resulting in a hydroxyl coverage of approximately one-third to one-half ML. The humid air can be air with a water content essentially at any value greater than approximately 5% and less than 100% and preferably a humidity value between 20 and 80%, assuming ambient pressure. The humidity should be at a sufficiently high value for approximately a ML of physi-sorbed water molecules to form. The copper was introduced into the vacuum chamber by magnetron source (other methods could include sputtering a Cu target or simple evaporation) thus depositing Cu onto the surface at room temperature. X-ray photoelectron and Auger electron spectroscopy (XPS and AES, respectively) measurements indicated that layer-by-layer growth (that is, wetting) occurred. The initial fractional Cu adlayer was oxidized to Cu(l) (
Further Cu deposition produced a metallic layer bound to the Cu ions (as illustrated in
Experiments were similarly run on a sapphire sample with decreasing amounts of surface hydroxylation. Dehydroxylation was accomplished by argon ion bombardment, followed by annealing in a partial pressure of oxygen to restore surface quality (smoothness) and stoichiometry. Results demonstrated that the degree of hydroxyl surface coverage was critical in the capability of copper to wet sapphire, with the degree of wetting decreasing as the dehydroxylation increased (as illustrated for a non-hydroxylated oxide surface in
The method of the chemical reaction (1) has been experimentally observed by the deposition of Co on hydroxylated alumina (
The invention being thus described, it will be obvious that the same can be varied in many ways. Such variations are not to be regarded as a departure from the spirit and scope of the invention, and all such modifications as would be obvious to one skilled in the art are intended to be included within the scope of the following claims. It is also obvious that at higher than normal microelectronic manufacturing pressures, steps will have to be taken to prevent the surface negative species from attracting and binding contaminants from the environment. Such steps can include the simultaneous deposition of the metal with the material containing the negative species.